Problem with AGP2PCI and Voodoo5 6000

[vykupitel]

Hello,

I found that this configuration is working only in "single chip" mode. Any other mode will sleep my monitor. I can run everything in 640x480 resolution but any higher will sleep monitor. Interesting is that problem depends on drivers too. On AM3.0 my monitor sleep but on 1.07 drivers it is working but the sreen is corrupted.

Does anybody here have any experience with that? Is there solution for that?

I have searched this forum and found that COSWORTH6K run this. So COSWORTH6K, could you help me? It is working for you?

[PanoramixDruida]

I really don't know anything about this. But you could try looking into http://www.voodoo6000.com/ . I guess there're skilled people at this site that could help you.

[vykupitel]

I found weird problem, it really depends on resolution,because it works in 1152x864 and higher,but on very low performance(fps around 4-5 in Q3). I depends somehow on selected FSAA mode.On higher FSAA coruption is  more frequent.

[r21vo]

did you use V5-6000 external power supply? (maybe agp2pci card somehow can't handle the high power requirements of VooDoo5-6000)

[vykupitel]

Yes, I have an internal hookup and AGP2PCI is configured for PCI power. I have tryed extrenal power for AGP2PCI but with no success.

[r21vo]

maybe try different agp2pci card if you have a chance to get one somewhere.

[vykupitel]

This problem have Hank Semenec too with his AGP2PCI.
I think that problem is related to SLI.

[r21vo]

but isn't here anyone who have used VooDoo5 6000 AGP with AGP2PCI adapter? Or at least VooDoo5 5500?

[vykupitel]

OK, I found that problem exist only in NT-kernel OS. It is working in W98SE. So I need workaround in driver. I could do any tests. Would be there any developer who would help me with it?

SSFT or Koolsmoky?

[koolsmoky]

Alright, since I added 6K SLI support to the NT drivers, I'll look into it. I'm also dealing with Q3D Obsidian2 and NT so don't *needle* me this time if it takes a while.

-KoolSmoky

[vykupitel]

Now I'm really sure that problem is in W2k drivers. They cannot find remain chips on the second PCI bus. Hank Semenec found this. In W9x is V56k working flawlessly.
In this case we will need some work. I hope that SSFT or Koolsmoky maybe will have some time.